Gerda

Technical Proposal

Version 0.1

The Gerda Collaboration

March 2005

Gerda

Technical Proposal V0.1

I. Abt j, M. Altmann j, A.M. Bakalyarov i, I. Barabanov g, C. Bauer c,M. Bauer l, E. Bellotti f , S. Belogurov g,h, S.T. Belyaev i, A. Bettini k,

L. Bezrukov g, V. Brudanin b, C. Buttner j, V.P. Bolotsky h, A. Caldwell j,C. Cattadori a,f , M.V. Chirchenko i, O. Chkvorets c, H. Clement l,E. Demidova h, A. Di Vacri a, J. Eberth d, V. Egorov b, E. Farnea k,

A. Gangapshev g, G.Y. Grigoriev i, V. Gurentsov g, K. Gusev b, W. Hampel c,G. Heusser c, W. Hofmann c, L.V. Inzhechik i, J. Jochum l, M. Junker a,

S. Katulina b, J. Kiko c, I.V. Kirpichnikov h, A. Klimenko b,g, K.T. Knopfle c,O. Kochetov b, V.N. Kornoukhov g,h, R. Kotthaus j, V. Kusminov g,

M. Laubenstein a, V.I. Lebedev i, X. Liu j, H.-G. Moser j, I. Nemchenok b,L. Pandola a, P. Peiffer c, R.H. Richter j, K. Rottler l, C. Rossi Alvarez k,

V. Sandukovsky b, S. Schonert c, S. Scholl l, J. Schreiner c,B. Schwingenheuer c, H. Simgen c, A. Smolnikov b,g, A.V. Tikhomirov i,C. Tomei a, C.A. Ur k, A.A. Vasenko h, S. Vasiliev b,g, D. Weißhaar d,M. Wojcik e, E. Yanovich g, J. Yurkowski b, S.V. Zhukov i, G. Zuzelc

a INFN Laboratori Nazionali del Gran Sasso, Assergi, Italyb Joint Institute for Nuclear Research, Dubna, Russia

c Max-Planck-Institut fur Kernphysik, Heidelberg, Germanyd Institut fur Kernphysik, Universitat Koln, Germany

e Jagiellonian University, Krakow, Polandf Universita di Milano Bicocca e INFN Milano, Milano, Italy

g Institute for Nuclear Research of the Russian Academy of Sciences, Moscow, Russiah Institute for Theoretical and Experimental Physics, Moscow, Russia

i Russian Research Center Kurchatov Institute, Moscow, Russiaj Max-Planck-Institut fur Physik, Munchen, Germany

k Dipartimento di Fisica dell’Universita di Padova e INFN Padova, Padova, Italyl Physikalisches Institut, Universitat Tubingen, Germany

Spokesperson: S. Schonert( Stefan.Schoenert@mpi-hd.mpg.de )

Co-Spokesperson: C. Cattadori( Carla.Cattadori@lngs.infn.it )

Technical Coordinator: K.T. Knopfle( ktkno@mpi-hd.mpg.de )

2

Contents

1 Preface to Version 0.1 5

2 The Sites of the GERDA Experiment in LNGS 6

2.1 Main Experimental Site of GERDA . . . . . . . . . . . . . . . . . . . . . . 6

2.2 Liquid Nitrogen Storage and Auxiliary Equipment . . . . . . . . . . . . . . 9

2.3 Space in Overground Laboratory . . . . . . . . . . . . . . . . . . . . . . . 9

3 Water Tank 14

3.1 Layout and Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.2 Water Plant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.3 Shielding Materials in Water Tank . . . . . . . . . . . . . . . . . . . . . . 15

4 Cryostat 18

4.1 Design Considerations and Specifications . . . . . . . . . . . . . . . . . . . 18

4.2 Engineering Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

4.3 Backup Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4.4 Multilayer Insulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.5 Materials and Production . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

4.6 Thermal Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

4.7 Integration of Cryostat within GERDA . . . . . . . . . . . . . . . . . . . . 28

4.8 Instrumentation and Auxiliary Equipment . . . . . . . . . . . . . . . . . . 29

4.9 Cryogenic Infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

5 Penthouse 34

5.1 Clean-Room Functionality and Radon Reduction . . . . . . . . . . . . . . 35

5.2 Mechanical Decoupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

5.3 Clean-Room Components . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

5.4 Electronics-Room Components . . . . . . . . . . . . . . . . . . . . . . . . . 38

5.5 Safety Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

5.6 Summary of External Requirements . . . . . . . . . . . . . . . . . . . . . . 39

6 Safety Aspects 40

6.1 Cryostat-Water Vessel System . . . . . . . . . . . . . . . . . . . . . . . . . 40

6.1.1 Leakage of Water Vessel . . . . . . . . . . . . . . . . . . . . . . . . 40

6.1.2 Leakage of Cryostat . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

6.1.3 Safety Valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

6.1.4 Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

7 GERDA Assembly at Lngs 44

3

8 The LArGe-facility 468.1 LArGe infrastructures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468.2 The LArGe system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488.3 Time schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

4

1 Preface to Version 0.1

The Gerda experiment [Ger 04] searches for neutrinoless double beta decay of 76Ge. Ger-manium diodes made out of isotopically enriched material are suspended in a superinsu-lated copper cryostat filled with liquid nitrogen (LN) or liquid argon (LAr). Because of itsradiopurity the liquid does not contribute to the background and serves as a shield againstexternal radioactivity. The shielding is completed by an outer layer of water, i.e. the cryo-stat is mounted in a water vessel. A clean room on top of the water vessel will house anair-tight lock. The latter is part of the cryostat volume and houses all mechanics for thediode suspension.

This technical proposal for Gerda serves several purposes. First, it contains the avail-able information relevant for the safety review at Lngs: the current design of the cryogenicvessel, the cryogenic infrastructure, and the water vessel. The latter also includes the sup-port structure for the above mentioned clean room and lock.

In a later stage, a complete report will describe the design of the entire experiment andwill then serve as a reference for the built detector.

This document will therefore become available in several iterations and different chap-ters will be added once the corresponding design is finalized. This approach was chosento allow for a timely construction of the experiment and to facilitate the safety discussionwith the Lngs laboratory.

The technical drawings of components of the Gerda experiment in this version of theTechnical Proposal are preliminary.

5

2 The Sites of the GERDA Experiment in LNGS

The Main Experimental Site of the Gerda experiment is located in Hall A of theLngs underground laboratory in front of LVD. The LN2 Storage Area and an AuxiliarySystem Area are both located in the TIR tunnel section between Hall A and Hall B. Inaddition, Gerda uses the Gerda LArGe facility for detector storage and refurbishmentin Phase I. This facility is located in the Interferometer Tunnel close to LUNA II(exLENS). All these Gerda sites are indicated in Fig. 1. Additional laboratory space is

Figure 1: The sites of the Gerda experiment in the Lngs underground laboratory.

needed aboveground for the control of the experiment, detector tests and material storage.

2.1 Main Experimental Site of GERDA

The layout of the main experimental site of GERDA is shown in Figs. 2 and 3 - see alsoFigs. 4 to 7 for more details and dimensions. The water vessel containing the cryostatwith the germanium detectors is placed close to the TIR tunnel at the maximum possibledistance from the LVD experiment. The so-called Super-Structure consists of a three floorlaboratory building between water vessel and LVD, and a frame extending over the watervessel providing thus the support structure for laboratory room above the cryostat.

The ground floor (7.80×4.60 m2) of the laboratory building contains the MachineRoom with vacuum and cryogenic equipment. Also the equipment for the water treatmentis located in this area.

6

Figure 2: Layout of the Main Experimental Site of Gerda in Hall A (dimensions in cm).

The first level (6.65 x 4.60 m2) hosts the Service Area of Gerda. In consists of theControl Room and laboratory space for on site repair and maintenance. In addition, spacefor a workplace of the Lvd experiment is foreseen on this floor.

The second level with the same area as the first level hosts the Detector Laboratorywhich is needed for the production, assembly, test and repair of the Gerda detectors.

The third level named ”Penthouse” extends over the whole water vessel. It hosts thelock through which the detectors are inserted into the cryostat, a clean room for detectorstring assembly and an electronics room for data acquisition. Its layout is described indetail in section 5.

A stair case incorporated in the Super-Structure provides on each level access to opengalleries which are located between Gerda and Lvd. These galleries of a width of 1.20 mare shared by Gerda and Lvd; they allow the access to the experimental areas of Gerdaand serve also as emergency exits for the Lvd experiment.

The Super-Structure is mechanically decoupled from the water vessel and the cryostat.The admitted load for all floors is 1000 kg/m2 while in the area of the inner lock (item 8in Fig. 18) a load of totally 6000 kg is foreseen. In this area, the support structure of thepenthouse keeps a clearance of Ø = 1.80 m for the neck of the cryostat.

7

An electrical hoist with a maximum load of 1500 kg is mounted on the third level.In addition all levels have docking positions which allow to bring in bulky equipment likefurnitures with a mobile elevator (see Fig. 3). A special lift (max. 200 kg) will connect thePhase II Detector Laboratory and the clean room areas of the Penthouse.

Figure 3: Lateral view of the GERDA experiment in Hall A (dimensions in cm).

As can be seen from Fig. 4 the dimension of Gerda allows for the passage of the cranefrom the north to the south of Hall A. However the hook of the crane must be moved toa lateral ‘east’ position. Safety systems have to be installed and operational procedureshave to be implemented in order to avoid collisions between the hook and the Gerdaexperiment.

All environments are air conditioned. The air conditioning of the Machine Roomand the Service Area guarantees a standard working atmosphere. The Penthouse andthe Detector Laboratory are designed as clean rooms. Thus a separate air treatment isnecessary for these rooms. The filter sets and and fan coils of both systems are located ontop of the penthouse.

Electrical Power is available on all floors. Power is needed on normal and no breaklines and at 230 and 380 VAC. The Gerda collaboration must still decide details on thetype of electrical power needed in the different environments of the Super-Structure.

Cooling water is needed for the cooling of the air conditioning systems. In addition,direct water cooling of some electronics devices is considered.

Washbasins are available in the Detector Laboratory and in the Penthouse area. Thewast water of these basins is collected in transportable containers of 1 m3 volume which are

8

located on the ground floor. These containers are exchanged by an authorised companyabout twice per month.

Sensors for fire, temperature and oxygen loss must be provided in all rooms.These sensors must be inserted in the LNGS supervision system. It should be notedthat these basic safety instrumentation is not to be confused with the GERDA Slow ControlSystem which is designed to monitor the status of the experimental apparatus and therelated processes. This system will be described in a future section.

Oxygen masks are available in all working environments and in the staircase. A spe-cific procedure will be produced in order to guarantee that every person working in thepenthouse has an oxygen mask on site.

An exhaust line for blow off of nitrogen gas generated during the normal operationbrings the nitrogen to the exit of the Underground Laboratory. This duct also takes theexhaust of the chemical clean benchs in the Penthouse and in the Detector Laboratory.

It may be possible/necessary to use this line also in the exceptional case of emergencyblow off. In this case the dimensions of the line have to be dimensioned accordingly.

2.2 Liquid Nitrogen Storage and Auxiliary Equipment

The LN Storage Area for Gerda is indicated in Fig. 1. Two 4 m3 dewars entirelymade out of stainless steel are mounted here. Superisolated lines connect these dewars tothe main experimental site of the Gerda. The consumption of LN of Gerda is estimatedto be of about 4 m3 per week. This consumption however depends on the details of thedesign of the cryostat and the related infrastructure.

The Main Electrical Switchboard of the experiment and part of the cryogenic equip-ment, which cannot be placed in the machine room of the laboratory building is locatedoutside Hall A as in Fig 1.

2.3 Space in Overground Laboratory

The complete running of the experiment is controlled from a dedicated External ControlRoom in the outside Lngs laboratory. Here, also the status of the experiment is monitoredcontinuously.

Experimental activities like testing photomultipliers and testing electronics are per-formed in outside Lngs laboratories in a dedicated External Experimental Areas.

During the construction phase material which arrives at LNGS is stored outside theUnderground Laboratory until really needed. Only exception are materials which mustnot be activated by cosmic rays.

9

Figure 4: View of Gerda cross section from TIR tunnel. The shielding structure belowthe roof of the water tank might be not needed.

10

Figure 5: Plan of water vessel and laboratory building

11

Figure 6: Plan of water vessel and laboratory building

12

Figure 7: Plan of water vessel and laboratory building

13

3 Water Tank

The Gerda water tank provides the outer shielding shell for the Gerda experiment whichhas the thickness of 3 m. The vessel houses in its center the cryostat in which the Ge diodesare operated. The water serves in addition as Cherenkov medium for the muon veto system.

3.1 Layout and Specifications

The main features of the Gerda water tank are summarized in Table 1 and a cross sectionis shown in Fig. 8, see also Figs. 4 to 7, and Fig. 9. The water tank consists of a cylinder

Figure 8: Cros section of watertank.

of 10 m diameter and 8.4 m height covered by a frustum-shaped roof which extends upto 8.9 m; the water level is at 8.4 m. The roof has a hole of 1200 mm diameter throughwhich the cryostat’s neck can be connected with the lock located in the penthouse abovethe water tank. The gap between neck and the roof must be closed by a flexible membranesuch that neither Radon nor light can enter the water volume. Further Radon rejection isachived by a slightly over-pressurized nitrogen blanket between water and roof. To read

14

out the Cerenkov light produced by the muons crossing the water, a structure holdingabout 80 photomultipliers will be fixed at the inner side of the water tank’s shell. Foroptimum light collection the walls of cryostat and water tank have to be covered with areflecting foil. Access into the water tank is possible by a manhole of 5 m diameter.

The water tank will be projected following the API 650 regulation, and the Eurocodice 8for the seismic acceleration, and built following the italian safety regulations (DL 626 andfollowing). The water tank construction is planned to start in autumn 2005, from a qualifiedcompany that will win the dedicated tender procedure. The Borexino water tank is areference concerning the safety aspects and the quality of the final product.

Due to its dimensions the water tank will be built on site. The procurement of materialand the needed technical gases will be procured by the company. To build the water tankthe needed equipment in Hall A will be:

• the 40 t crane of Hall A, and another crane provided by the collaboration will be nec-essary to install and keep in position the inner cryostat while the WT is constructed.In fact due to their geometrical dimensions the WT will grow around the cryostat.

• electrical power .?. kWh

• water (inlet and outlet)

3.2 Water Plant

The Borexino water plant will be used for producing the about 650 m3 of purified waterneeded to fill the Gerda water tank. The water will be transported by a pipeline fromBorexino to Gerda as shown in Fig. ?tobeprepared?.

To maintain the quality of the water in the Gerda water tank continuous recirculationof the water at about 2 to 3 m3/h and adequate water processing is foreseen. Details ofthis water plant are to be specified.

3.3 Shielding Materials in Water Tank

The limited height of the water tank implies insufficient vertical shielding which will becompensated by layers of radiopure copper or lead. These layers have to be installed belowthe cryostat at the bottom of the water tank and - depending of the final position of thecryostat within the water tank - also above the cryostat hanging there from the roof of thewater tank. (see ‘wings’ in Fig. 9).

The optimization of these shielding layers is done by Monte Carlo simulations.

15

Table 1: GERDA Water Tank main features

Reference regulation for structural project: API650Further verification for seismic hazards: Eurocodice 8Quality certification of construction process: ISO9001Quality certification required for company: ISO14001Tank height / external diameter: 8.9 m / Ø10.0 mHeight of the water level: 8.4 mEffective capacity (m3): 633 m3

Water tank bottom: flat, plates head weldedWater tank roof: conical from the shell, (Ø 4.5 m)Water tank shell: cylindrical, plates head weldedWater tank sheet-metal plates: ≈2 mAngle between shell and roof: ≈6◦

Bottom renforcement: yes, at 1 feet levelRenforcement rings along the shell: yes, 1 or 2Water tank Material: stainless steel AISI 304 L or 304 LN, or

carbon steel plus appropriate coatingThickness of the shell: 12 to 9 mmConnections between plates: weldedWelding type: external MIG, internal TIG without

filler metalWelding certification: certified by the executing company

fully X-ray testedApproximative length of welds: 400 mFlanges1 DN 5000 with davit lateral at bottom1 DN 1800 in roof for cryostat neck6 DN 500 in roof for level, pressure and

temperature sensors incl. spares1 DN 300 for total drain compensation2 DN 250 in roof for photomultiplier cables1 DN 200 to drain tank completely in 20 h2 DN 32 lateral for water recirculationWeight of water tank (tons): <20 tonsWeight of filled water tank: 650 tonsOperational over/underpressure: ±20-30 mbarSafety device: pressure/depressure ±20-30 mbarWater recirculation: yes, 2-3 m3/hWater recirculation plant: deionization, Radon stripping,

particulate removal

16

Figure 9: View from the TIR tunnel onto super-structure and water tank with cryostat.The shielding structure below the roof of the water tank might be not needed.

17

4 Cryostat

The GERDA cryostat serves as container for the liquid nitrogen (LN) or argon (LAr) inwhich the Ge diodes are operated. The cryogenic liquid serves simultaneously as a shieldagainst the remnants of the external γ background penetrating the surrounding water shieldand of the cryostat’s own radioactivity. In order to prevent any radon contamination viathe air, the cryostat will be operated at a slight overpressure of 49 kPa (490 mbar). Athermal shield system reduces the heat load from the surrounding environment (water)such that less than 0.2% cryogenic liquid are evaporated per day.

4.1 Design Considerations and Specifications

The size of the cryostat must allow to transport the complete vessel from its manufacturingplant to the Lngs underground laboratory by road. This constraint fixes its diameter tonot much more than 4 m. Using high performance superinsulation as thermal shield whichneeds very little space, about 2 cm, the radius of the LN shield will amount to about 1.9 mwhich implies severe constraints for the radiopurity of the building materials. Thus, inorder to achieve the desired background index of 10−3 cts/(keV·kg·y), only selected copper(or lead) is known to exhibit the required 228Th activity of less than 25 µBq/kg. Thecryostats minimum height amounts to 6-7 m, i.e. 4 m plus the space for the Ge-diodearrangement (0.5 m) plus the space for a neck (1.5 m) serving both as thermal impedanceand port for insertion of the diodes. Up to 1 m additional height of cryogenic liquid willhelp to compensate the loss of shielding material due to the hole in the neck. The supportof the cryostat will add at least 0.5 m in height to guarantee an efficient water shieldagainst neutrons; however, a thicker water layer at bottom, up to 3 m like reaially, wouldbe economic since else lead or copper shileds are needed for compensation.The following list shows a summary of the general specifications:• design and manufacture conforming to AD2000 Regelwerk [AD 2000] for pressure

vessels, DGRL 97/23 EG, and other applicable codes,• maximum outer radius of vessel system not more than 2.0 m,• thickness of cryogenic liquid, LN or LAr, in radial direction at least 1.85 m,• minimum height of cryogenic liquid in axial direction 5.2 m,• operating pressure 49 kPa,• use of low-radioactivity copper (25 µBq/kg 228Th) as major construction material,• other used construction materials to be shielded accordingly,• use of metal gaskets,• redundant instrumentation conforming to AD2000 and other applicable codes,• loss of cryogenic liquid by evaporation less than 0.2% per day,• earth quake resistance up to 0.6 g horizontal and vertical,• operating in water vessel of 9 m height,• operating temperature 20◦C, lifetime >10 years.

In addition the cryostat will be equipped with a cooling system that allows to reduce theevaporation rate of the cryoliquid by use of a heat exchanger.

18

4.2 Engineering Description

Fig. 10 shows the cross section of the superinsulated cryostat designed in accordance to theGerman codes for cryogenic and pressure vessels using copper of the DIN type DHP R240(formerly SF-Cu, material # CW024A, formerly 2.0090); the detailed production drawingis shown in Fig. 11. A summary of its characteristics is given in Table 2. The cryostatconsists of two coaxial vessels which are built from hemispherical shells of inner radii of1.89 m resp. 2.00 m and corresponding cylindrical centerpieces of about 1 m height. Bothvessels have a coaxial cylindrical neck of 1.5 m length; a major part of it is made fromstainless steel (1.4571) for achieving a high thermal impedance. The upper ends of thenecks are connected to each other so that the inner vessel is suspended via its neck in theouter vessel without any further thermal contact. Additionally, the vessels are centered atbottom by a cylinder of about 30 cm diameter and 16.5 cm height and a correspondingsocket that are welded to the inner and outer vessels, respectively. A cylindrical skirt madefrom 8 mm thick copper supports the outer vessel from the ground.

The volume between both vessels will be pumped via two about 2 m long pipes inorder to maintain the insulation vacuum below 0.01 Pa. Four further flanges will be usedto connect safety and pressure relieve valves as well as manometers.

The pressure vessel code requires a maximum thickness of the DHP R240 copper platesof 15 mm. This constraint implies that the earth quake tolerance of 0.6 g can be fulfilledfor horizotal accelerations only; in vertical direction 0.5 g can be tolerated. The crticialpart is the transition from neck to hemisphere, and thus the neck is reinforced by twostiffening rings which, however, degrade thermal insulation. Work is in progress to solvethis problem by manufacturing the complete neck including an adaptor ring from stainlesssteel.

An alternative design is shown in Fig. 12, see also Table 2. The crucial difference withrespect to the former design is that now the inner vessel rests on 12 glass-fiber reinforcedplastic (GPR) pads which are located at bottom between inner and outer vessel. Thecompensation for thermal shrinkage of the inner container, a maximum of 36 mm whencooled down from 300 K to 77 K, is provided by a bellow in its neck. Three spacers withinthe neck and three spacer at bottom keep the inner vessel centered. Finite element analysishas proven it to meet all earth quake requirements despite of an increase in the height ofthe centerpiece to 2 m.

A final decision between both options will be made as soon as both design are finalizedand more information about the GRP material and the copper-stainless steel weld areavailable.

19

Figure 10: Preliminary design for a copper cryostat with the inner container hanging fromthe neck.

20

Figure 11: Production drawing for the preliminary version of a copper cryostat with theinner container hanging at neck.

21

Table 2: Preliminary technical data for two copper vessel options.

Design option 1 option 2

Materialsvessel bodies Cu-DHP R240 dtoradiopurity 232Th <30 µBq/kg dtoradiopurity 238U dtoneck all stainless steel copper + ss bellowpads none GRPGeometryoverall dimensions Øo×h 4034×7613 4000×9100 [mm2]outer vessel Øo×h 4020×5083 4000×6000 [mm2]inner vessel Øi×h 3784×4600 [mm2]neck height 1500 1600 [mm]neck inner diameter 800 dto [mm]inner vessel volume incl. neck (41) (52) [m3]volume for liquid gas (33) [m3]Massesempty vessel 17,7 [103kg]LN fill 26,6 [103kg]LAr fill 46,0 [103kg]Pb shield [103kg]total max. mass 63,7 [103kg]Pressuresinner vessel operating overpressure 0.5 dto [105 Pa]inner vessel maximum pressure 2.5/-1 dto [105 Pa]pressure for LN / LAr emptying [105 Pa]outer vessel overpressure -2 dto [105 Pa]Superinsulationnumber of superlayers 10 - 30 dtonominal insulation vacuum 10−2 dto [Pa]leakage rate 10−5 dto [m3Pa/s]thermal loss (calc./upper limit) / [W]corresponding daily loss of LN [`]Safetyconstruction code AD 2000 HP dto

DGRL 97/23 EG dtoearth quake tolerance h/v : 0.6/0.5 g 0.6 g

22

Figure 12: Preliminary design for a copper cryostat with the inner container resting onepoxy pads.

23

4.3 Backup Solution

Until the feasibility of a copper cryostat is clearly established work continues on an al-ternative solution i.e. a stainless steel cryostat that carries a Cu or Pb shield in the coldvolume (see also section 6.2.3 of the Proposal). Fig. 13 shows the most recent design witha highly optimized graded internal copper shield. The dimensions of the vessel are nowvery much compatible with those of the copper cryostat and the weight of the shield islargely reduced compared to the version shown in the Proposal.

Figure 13: Preliminary design for a stainless steel cryostat with internal copper shield.

24

4.4 Multilayer Insulation

The multilayer insulation (MLI) will be attached in the space between both vessels to thewall of the inner vessel. If the inner vessel is resting on GRP pads, the MLI has to extendover the sides of these pads. The maximum thickness of the MLI will not exceed 20 mmincluding assembly and attachment provisions.

MLI or superinsulation consists of closely spaced shields of Mylar or Kapton coveredwith a thin reflective layer of aluminium, silver or gold. Stacks of typically 10 layers withintegrated spacers are tailored by industry for specific shapes and applications; dependingon material, number of layers, and packaging density the thermal flux ranges from 0.55 to1.5 W/m2, see Table 3 for one specific product line.

Table 3: Specifications of several commercial MLI thermal blankets for temperatures of77 K and 300 K [AAe 05]. All products contain polyester foils or yarns.

Product heat loss density

W / m2 g / m2

Coolcat 1 crinkledone side aluminisation, venting at edges30 foil layers 1.0 - 1.5 234 - 27640 foil layers 0.8 - 1.250 foil layers 0.6 - 0.9Coolcat 2 spacered - meshdouble side aluminisation, perforated10 foil + 10 space layers 1.0 - 1.5 128 - 182 or 207 - 27720 foil + 20 space layers 0.8 - 0.930 foil + 30 space layers 0.7 - 0.740 foil + 40 space layers 0.55 - 0.6Coolcat 2NW spacered - porous structuredouble side aluminisation, perforated20 foil + 20 space layers 0.79 256 - 364 or 414 - 55440 foil + 40 space layers 0.65

To function as a high performance thermal insulation MLI reqires high vacuum below0.01 Pa. Fig. 14 shows the variation of heat flux with vacuum pressure for another com-mercial product consisting of 30 layers with 7.0 mm total thickness. Above 1 kPa the heatflux has increased by more than two orders of magnitude. Thus, the vacuum pressure willbe monitored and maintained by periodical pumping.

To avoid degraded vacuum conditions between the foils of large area MLI blanketsystems, the purely edge-pumped blankets of the past have been replaced recently byperforated blanket systems. These exhibit for example non-overlapping holes of typical4 mm diameter and a pitch of 120×156 mm yielding 0.05 to 0.1% open area.

25

Figure 14: Variation of heat flux (right scale) and apparent thermal conductivity in a MLIblanket (30 layers, 7 mm total thickness) with cold vacuum pressure [Fes 02].

In case of polyester insulation suitable welding protection is required since this materialis flammable. Polyester foil is known to fail in test 4.4.2 of EN 1797: flash in drop testunder 50% LOX - 50% LN [AAe 05].

4.5 Materials and Production

The required radiopurity of the copper has been verified so far for the copper grade NOSVonly which had been delivered by the Norddeutsche Affinerie AG to the MPI Heidelberg.The required DHP copper grade is a phosphorus-deoxidized product; it can be producedfrom the NOSV grade by being alloyed with 0.015 to 0.040 (mass)% of phosphorus.Technically, a phosphorus-copper granulate with 10 (mass)% of P is added in the electricallyheated melting furnace to the NOSV material. Thus the radiopurity of the resultingDHP copper can be deduced most easily from the radiopurity of the phosphorus-coppergranulate. A preliminary value of its 228Th activity is (10±2) mBq/kg; hence the additionof 3 g P-Cu granulate to 1 kg of NOSV grade copper will increase its radioactivity by35 µBq/kg which is a factor of 1.5 higher than specified. Work is in progress to reduce the228Th content of the granulate.

The radiopurity of one type of thermal insulation, NAC-2 cryo-laminated superinsula-tion foil with bonded space material according to HERA/DESY specifications, has beenmeasured to be less than 50 mBq/kg (upper limit). Assuming 20 layers of a density of24.2 g/m2, this would increase the radioactivity budget by less than 140 µBq/kg. Thus, amore sensitive measurement is needed.

The radioactive contribution from all other construction materials, stainless steel andGRP pads, is known to be much higher; typical values are 10 mBq/kg and the amount

26

of material used locally is in the order of kg at least. Thus shielding of these materi-als is needed, and the designs of the cryostat account for that by including additionalloads of 1000 kg at the transition neck-hemisphere and at the bottom of the inner vessel,respectively.

To prevent contamination by additional radioactivity, the welding and assembly ofthe cryostat has to be done at clean room conditions. Also in this context, electronbeam welding is the optimum welding technique being carried out in vacuo and needingneither welding electrodes nor additives. A welding facility with a sufficiently large vacuumchamber of 6×7×14m3 has become available recently [Pro 04]. First electron beam weldingtests are in progress including the weldment of two pressed spherical segments from DHPcopper as well as material tests for joints of DHP copper and stainless steel.

4.6 Thermal Performance

The heat flow Q through a material of cross section A and thickness t is given by

Q = λ · A ·∆T/t

where λ is the material’s thermal conductivity and ∆T the temperature difference betweenits surfaces A. In the following ∆T is assumed to be (293-77) K = 216 K.

For a neck of 0.8 m diameter made out of 6 mm thick stainless steel (λ = 15 W/m·K)and a of 1.5 m length, A = 0.0151 m2, and hence Q = 33 W.

The heat flow through 12 pads of glass fiber re-enforced plastic (λ(GRP) = 0.3 W/m·K)of 0.052 m2 area and 0.1 m thickness amounts to 19.4 W.

The thermal conductivity of a 20 layer superinsulation is about 0.0023 W/m2K between300 K and 77 K. Assuming a cryostat’s surface area of 74 m2 and a safety factor of 2, theheat flow amounts to 74 W (see also Table 3).

Table 4 summarizes the thermal losses discussed above as well as the correspondingdaily evaporation rates for nitrogen; for Ar the evaporation rate is a factor of 1.23 larger.With a total cryogenic mass of 39000 (67000) kg of LN (LAr) the daily evaporation rateis for each option well below 0.2%.

Table 4: Thermal losses

Cryostat part W kg(N2)/d kg(Ar)/dNeck 33 14.3Bellow 10? 4.3GRP pads 19 8.2Superinsulation 74 32.0Sum(inner vessel hanging) 107 46.4 57.1Sum(inner vessel on pads) 103 44.7 55.0

27

4.7 Integration of Cryostat within GERDA

The cryostat will be mounted onto a vibration-isolated basement and will be completelydecoupled from the surrounding Super-Structure. If the cryostat cannot be grounded via asuitable grounding bar in (the floor of) Hall A, its mount must be also electrically isolatedfrom the floor.

The cryostat will be connected via a 80 cm long cylindrical interface of 80 cm diameterwith the lock (see sections 2 and 5). This manifold is made from stainless steel and needsto be thermally isolated. It contains the feed throughs for all devices that penetrate intothe cold volume. At present, the following tubings and feed throughs are foreseen (seeFig. 15).

• superinsulated filling tube

• gas exhaust tube

• superinsulated pipes for active cooling

• feed throughs for all instrumentation outlined below

• feed throughs for detector cables

• several spare lines

Detailed specifications will be presented in a future section.

Figure 15: Schematic of the manifold through which indicated devices penetrate into thecold volume.

The connections between cryostat, interface and lock exhibit metal-gaskets and mustpass a He leak test.

Cryostat, manifold and lock must have also a low impedance electrical connection.

28

4.8 Instrumentation and Auxiliary Equipment

The instrumentation and auxiliary equipment allows to monitor and control the status ofthe cryostat such that a reliable and safe operation is guaranteed. Devices that are crucialfor safety will be redunantly implemented.

The instrumentation for monitoring and controlling the cold volume includes ( Fig. 15)

• level indicator of cryoliquid (hydrostatic and/or capacitive device)

• level indicator for cryoliquid (high resolution)

• manometers (2×)

• pressure control valve

• overpressure safety valve and rupture disk

• underpressure safety valve and rupture disk

• thermometers (several)

• slow control system

The evaporated cryoliquid is pressure-less substituted by liquid stored in a small superiso-lated container of about 1 m3 that is located close to the top of the cryostat’s neck. Thiscontainer, in turn, is refilled via a pressurized line from the big storage tanks in the TIRtunnel.

The instrumentation and infrastructure for controlling and maintaining the insulationvacuum includes

• manometers (2×)

• mass spectrometer for identifying type of leak

• overpressure safety valve and rupture disk

• roughing and venting valve

• diaphragmas for roughing and venting

• slow control system

The insulation vacuum is served by one or two turbopumps which are located close to thetop of the neck. The roughing pump(s) could be installed in the machine room of thelaboratory building. Pump-down and venting must be done in a well controlled way sothat the MLI is not damaged. For that purpose, the conductance of the line for roughingresp. venting will be controlled by diaphragms of appropriate size. In normal operation,the turbopumps will be shut off and the valves between pumps and cryostat flanges willbe closed.

29

4.9 Cryogenic Infrastructure

The cryogenic infrastructure includes

• One storage tank for liquid nitrogen (LN) and one for liquid argon (LAr). The latterwill be needed for the operation of the cryostat with argon. These standard vesselswill have a volume of 4 m3 each and will be located in the TIR tunnel (see chapter 2).The cryoliquids will be supplied to the Gerda experiment by superisolated pipes.

• An active cooling unit to avoid / reduce evaporation losses during operation. About0.2% of the cryostat volume will evaporate per day. This corresponds to a heatingof about 200 W.

Just below the fill level in the cryostat a heat exchanger (copper tubes) will beinstalled and liquid nitrogen with lower temperature at sub-atmospheric pressurewill circulate inside. Part of the equipment to provide the coolant will be installedon the gallery of the 3rd floor.

• Equipment for initial filling and emptying of the vessel. Different options exist fordraining of the vessel: overpressure in the gas volume, pumping of the liquid, orevaporation with a heater. Since it is not anticipated to exchange the cryogenicliquid during the lifetime of the experiment - except if nitrogen is exchanged byargon - the last option might be a viable one. As an example, the drainage will take2-3 weeks with a 5 kW heater. In case overpressure is used, about 2 bar (absolute)is needed and hence the vessel will be classified as a pressure vessel according toEuropean rules.

• A radon cleaning unit for the removal of radon from the liquid nitrogen/argon. Thisunit may only be needed during the first filling of the tank. A final decision can onlybe made once all parameters of the active cooling are known. Such a unit is alreadyin operation for Borexino and the Genius Test Facility. It would also be placed atthe ground floor of the laboratory building.

The overall space required for the equipment at the ground floor of the laboratorybuilding is estimated to be about 20 m2. The electrical power for pumps and control unitsis expected to be 5-10 kW.

A final design for the cryogenic infrastructure is not yet available but the basic featuresare known and two competing options are available. Both include drawings for the layoutof the infrastructure together with a description of the operation. Some iterations areneeded needed to come to a final design.

Fig. 16 shows the conceptual design by CryogenMash. The coolant for the heat ex-changer (item 8) is liquid nitrogen taken out of a storage tank operated at an absolutepressure of 2 bar. The nitrogen is expanded through a nozzle. The resulting liquid/gasmixture will boil in the heat exchanger and the gas is then pumped by a vacuum pump(item 6). The pressure (temperature) of the cryostat is measured by the manometer P1(temperature transmitter T1) and it is controlled by the flow through the pump (valve

30

VE2) and by the valve VE3. The fill level is measured by the hydrostatic pressure of thefluid (gauges L1 and L2). All lines with liquid nitrogen are routed from the cryostat tothe ground floor of the laboratory building through one vacuum insulated pipe of DN 300.In the schematic, several pipes are inside the insulation vacuum of the cryostat and theinlet/outlet is at the bottom of the tank. For safety considerations, this solution will notbe chosen, and we will avoid any openings in the vessel below its fill level. Instead, therewill be one copper pipe inside the cryostat volume which can be used as inlet and forthe hydrostatic pressure measurement. The passive safety valves S1/S2 open in case thepressure regulation fails. The insulation vacuum of the cryostat is pumped by a cryogenicgetter pump (item 5). A heater (item 4) of 15 kW is able to evaporate some nitrogen andthe resulting gas can be used to overpressure the vessel and hence to empty it through thevalve V3.

An alternative design for the refilling and cooling of the cryostat by Sommer Labortech-nik is shown in Fig. 17. The liquid nitrogen from a high pressure storage tank is expandedto a smaller tank of ' 1 m3 volume (left tank in figure) with an absolute pressure of 0.6 barin the gas phase. The underpressure is maintained by a turbo blower. The liquid nitrogentemperature is therefore cooled to about 73 K, and it will be used as a coolant for theheat exchanger inside the vessel. The advantage of this solution is that no boiling occursin the cryostat and thus microphonic noise is avoided during operation. In addition, thecooled nitrogen can be used for refilling. The 1.4 bar absolute pressure inside the cryostatis maintained and controlled by a commercial system (right tank in figure).

The control system for the electromagnetic valves and measurement devices will be aProgrammable Logic Control (PLC). One possible solution is the DeltaV system of Em-merson Inc. since it is already used by experiments like Borexino and expertise exists atLngs. The PLC and the safety control devices should be connected to an uninterruptablepower supply.

31

Figure 16: Schematic of the cryogenic infrastructure according to a design by Cryogenmash.

32

Figure 17: Schematic of the refilling and cooling system according to a design by SommerLabortechnik.

33

5 Penthouse

The penthouse is located on top of the experiment. It is the upper part of the super-structure also encompassing the platform above the water vessel and the laboratory build-ing on the back-side of the experiment. The super-structure as a whole is decoupled fromthe water vessel.

Functionally the penthouse is separated into a clean- and an electronics-room. Theclean-room houses the lock system which allows the insertion of detectors and calibra-tion sources. The lock itself is mechanically supported by the platform, but mechanicallydamped. The electronics-room contains the needed racks and support systems. Also in-corporated is a control area which allows commissioning and debugging.1 The penthouseor parts of it will be shielded as a Faraday cage according to specifications expected fromthe electronics group. Fig. 18 shows the layout of the penthouse.

Figure 18: Layout of the penthouse [int.vers. 8] on top of the vessel with clean-room, locksystem and the electronics-room. Numbered components are specified in subsection 5.3.

1This is not the control room for normal operation which is located on the first floor of the laboratorybuilding.

34

The footprint of the penthouse is determined by the platform and the laboratory build-ing. Due to requirements of the neighboring experiment [LVD] the penthouse has a cut-out.In this area a platform on top of the staircase services crane operation for heavy parts andaccess to the penthouse. The penthouse is surrounded by a gallery for access and safetypurposes.

The overall floor load is specified to 1 t/m2. Above the neck of the cryostat the innerlock incorporates a lead shield of approximately 2 t. The complete lock is estimated toweigh about 6 t. This load has to be supported by the platform.

Fig. 4 shows a vertical cut through the experiment looking from the street. The currentdesign tries to exploit the full height of hall A. This does not allow the crane to pass overthe experiment. Probably, a special crane will have to bring up loads to the cut-out area.

The penthouse ceiling has a dome above the lock in order to facilitate vertical lift. Themuon veto is partially placed on top of the dome, partially around it. It is assumed thatmost of the cryogenic support is installed on the left side (viewed from TIR) which is easierto access. The drawing shows two lines at 500 mm and 800 mm above the water tank.To the left of the lock all 800 mm are allocated to the cryogenics. On the right side achannel for cables running between the lock and the electronics room is foreseen betweenthe 500 mm and 800 mm levels. The electronics-room is lowered to the 500 mm level. Itshould be noted that the maximum rack height will be 1600 mm.

The final space requirements of the platform and the internal allocations will have tobe rediscussed when a design of the water-tank and the super-structure becomes available.The current space allocations allow a height of 8900 mm for the water tank. If additionalspace for the super-structure is needed, the height of the water tank will have to be reduced.

5.1 Clean-Room Functionality and Radon Reduction

The clean-room with its integrated lock system will provide the possibility to insert andwithdraw detectors in a modular way while the vessel stays at cryogenic temperatures. Theclean room will also be used for the final preparation of the detectors and their integrationinto the detector suspension system. The detectors are delivered either through a materiallock located in the cut-out of the penthouse or through an internal lift from the detectorlab on the second floor of the backside house.

Radon is a major concern for the experiment. The interior of the cryostat has to bebasically ”radon free”. The inner lock is connected to the main gas volume of the cryostat.Therefore the partial pressure of Radon in the clean-room has to be reduced in order tosuppress Radon creeping into the lock. The air of the clean-room will thus be filteredthrough a Radon reduction plant.

The second consideration is the attachment of Radon to aerosol particles which settleon the detectors. Radon decays into 210Pb which is a dangerous source of background onsurfaces of n-type detectors. The clean-room will be of class 10000 which is sufficient tosuppress this problem when combined with Radon reduction.

For some of the handling the flow boxes can be provided with synthetic air to furtherreduce the danger of contamination of the detectors. For storage facilities a pure nitrogen

35

atmosphere is foreseen.

Materials :

All wall and components will be constructed out of stainless steel. Any window neededwill be acrylic. All materials will be cataloged and tested for radon emanation

Environmental Control :

Particle counters will be available to check the condition of clean-room and work-stations. In addition the oxygen content and humidity will be controlled.

5.2 Mechanical Decoupling

The super-structure is one mechanical unit. It is decoupled from the water tank.

The lock system is mechanically decoupled from the supporting platform. This is tokeep the position of the detectors inside the vessel stable and reduce vibrations.

5.3 Clean-Room Components

The main components of the clean-room infrastructure can be located through their num-bers in fig. 18. They are:

1. Personnel-lock

2. Air-shower

3. Material-lock

4. Container loading station

5. Detector Preparation Tables

6. Final assembly station

7. Outer lock

8. Inner lock

9. Cable bridge

10. Detector storage unit

11. Tool-cabinet

36

1,2: Personnel enter the clean-room through a classical personnel lock and an air-shower. In the lock private storage space will be available.

3: Materials, especially detectors, can enter the clean-room through a material-locksuitable for containers up to l=60 cm×d=50 cm×h=40 cm. The lock can be evacuated,heated [80 ◦C] and flushed with nitrogen.

4,5,6,7: All these stations will be laminar flow-boxes. When not in use they are closedand flushed with nitrogen. As a result, detectors which are under preparation do not haveto be transferred if unpredicted breaks are necessary. All station specific tools are kept inholders within the station.

4: The material-lock opens towards the container loading station. A passive rollersystem is foreseen. Storage space for 4 containers will be provided under the work-bench.

5: This station will be used to prepare detectors and cables. All work that can be donebefore the integration of a detector into a package will be done here. Pressured clean airwill be available.

6,7,8: These stations are connected via a rail system below the ceiling of the locksystem. Detector packages are hung on movable hooks. Packages can be rotated. Thewindows between the stations are opened to allow the transfer from one station to thenext. The maximal package height is 150 cm.

6: Lowered tables with integrated suction systems allow easy access to the hangingpackages. Packages can be temporarily stored in the area close to the outer wall whenthe area is closed off. During the final assembly the detectors are integrated into so calledstrings and all cables are combined into easily pluggable units. Pressurized clean air ornitrogen will be available. While open the laminar flow consists of artificial air. Duringtransfer to the outer lock both stations are filled with nitrogen.

7: The outer lock always contains an ultra-pure nitrogen atmosphere. While closed offagainst the inner lock and the final assembly station it can be evacuated, heated [50 ◦C]and flushed. After the last cleaning cycle the outer lock will be flooded with nitrogen fromthe boil-off of the vessel and the connection to the inner lock will be opened. A mechanicalarm grabs the cables hanging inside the inner lock. After they are pulled into the outerlock they are connected with the help of a glove box integrated in the outer lock. Theglove box is shuttered off during non operation and constantly flushed. After the cablesare connected the package is transferred into the inner lock.

7,8: During normal operation part of the boil-off from the tank is transferred to theinner lock. Gas streams through a valve into the outer lock. The inner lock has thesame pressure as the vessel, 1.2 bar. The outer lock has a pressure of ≈1.15 bar, which isintermediate between the inner lock and the clean-room. During transfers from the outerto the inner lock both volumes are kept under 1.2 bar.

8: The neck of the vessel opens into the inner lock. During insertion operations it isbasically a part of the volume of the vessel. During normal operation the neck will beclosed and only part of the boil-off will be lead into the lock.

A positioning and support plate for the suspension system will be placed inside the neck.It is placed above the rupture disk or safety valves for cryogenic emergencies. Duringnormal operation the strings are hanging from this plate. If further MC calculations

37

indicate the necessity of further neutron shielding a polyethylene plug will be lowered intothe plug above the plate. A lead lid is incorporated at floor level. It is opened for insertionoperations. Additional polyethylene can be placed above the lead.

Cables enter through a panel from the cable-tunnel, see 9. They are fixed to holdersinside the lock.

The inner lock has a web-cam for constant surveillance.9: A cable tunnel between the lock and electronics room will minimize cable lengths.

It is integrated into the super-structure below the clean-room floor. Special feed-throughsbring the signals from the lock into the tunnel.

10: The detector storage unit allows the storage of detectors and other materials underradon reduced nitrogen atmosphere.

11: The tool cabinet holds all non station specific tools, especially the ones neededfor mechanical operations steered from outside the lock system. This cabinet has normalclean-room atmosphere.

5.4 Electronics-Room Components

The systems to be supported are

• signal processing and DAQ

• fast control

• high voltage

• slow control and low voltage

• environmental monitoring

• mechanical support

The racks for signal processing [FADC] have to be temperature controlled.

5.5 Safety Considerations

Clean-Room: The oxygen content of the clean-room atmosphere will be constantly mon-itored. Oxygen masks will be provided for accidents.

As the direct view into the hall is obscured, video monitoring of the hall will be availablein order to allow fast assessment of external dangers. The inner lock is video controlled toassess possible internal mechanical failures.

There are temperature and pressure sensors in all areas of the lock. There are mechan-ical safety valves to prevent over pressure in the lock system.

An internal alarm system as well as the connection to the Lngs general alarm systemsinform about all safety relevant failures.

Escape routes follow the main staircase or the connection between the penthouse andhall galleries.

38

Lock: The lead cover of the neck which is incorporated in the inner lock can be retractedwithin 5 to 10 minutes. However, in case of a sudden build-up of pressure in the cryostat thelead cover suggests to have the blast disk below the inner lock. The positioning plate of thesuspension system should also be located above the blast disk. Thus only the suspensioncables will be below the blast disk.

During normal operation part of the boil off from the cryostat is used to flush the innerand outer lock. In case of a malfunction the pressure could build up inside the lock system.Safety valves will be incorporated in order to prevent dangerous conditions.

5.6 Summary of External Requirements

Electricity: During operation the electronics will consume most of the power. Thestrongest power consumption is expected to come from heating systems in the lock areas.

• 240V:

– heating for inner/outer lock/material lock,

– electronics racks,

– computers etc.,

– motors,

– lighting [halogen],

– flow boxes,

– air conditioning clean-room/electronics room

• UPS:

– oxygen monitoring of clean-room

– video monitoring of out-side and lock area

– pressure and temperature monitoring of the lock area

– alarm system for penthouse [and laboratory building]

– some parts of electronics

– some parts of data acquisition system

Cooling Water: The electronics will be water cooled using a regulated closed watersystem. External cooling water will be used to cool the system.

Nitrogen: All gas supplies will come from bottles or boil off from the cryostat resp.storage tanks.

39

6 Safety Aspects

6.1 Cryostat-Water Vessel System

Figure 19 shows a schematic of the Gerda vessel system which is used to discuss some ofits safety aspects. Its characteristic feature is a cryostat being completely immersed into awater tank. The cryostat contains either LN or LAr; its volume is 46 m3 which correspondsto 32000 (39000) m3 of N2 (Ar) gas. The water volume is 660 m3. Safety-relevant eventsmight be due to the leakage of either the cryostat or the water vessel.

Figure 19: Vessel dimensions assumed for discussion of safety aspects. (1) indicates theneck of the vessel, (2) the tube through which the space between inner and outer containeris pumped. Within this space the multilayer insulation (MLI) is mounted.

6.1.1 Leakage of Water Vessel

While the water contained in the GERDA vessel is purified and, hence, cannot pollute theGran Sasso natural water, the sheer amount of water released through a major leak couldrepresent a problem for nearby installations. Only 350 m3 can be retained in the emergencypit. Thus, it might be recommendable to either build an appropriate water retaining bassinaround Gerda and/or to install a pipe for draining the water to the outside.

6.1.2 Leakage of Cryostat

The cryostat consists of two vessels which are connected at the neck. The space betweenboth containers is evacuated and - for improved thermal insulation - equipped with amultilayer superinsulation. Possible failures caused by the cryostat are

40

• loss of insulation vacuum,

• leak in outer vessel,

• leak in inner vessel,

• leaks in both vessels.

Loss of vacuum for superinsulation: The loss of the insulation vacuum will cause adegradation of the thermal insulation of the cryostat. Fig. 14 shows this degradation for aspecial case to be about two orders of magnitude, while the compilation in Fig. 20 indicateseven larger factors of up to 103. For a most conservative estimate, a superinsulation with

Figure 20: Thermal conductivity of superinsulation in dependence of gas pressure [Tim 89].

a heat flux of 1 W/m2 between 77 and 300 K is assumed to be degraded by a factor of1000 due to the loss of the insulation vacuum. With the cryostat’s surface of 63 m2 and37 tons of LN (evaporation enthalpy: 199.3 kJ/kg) it will take about 32.5 hours until allLN has been evaporated. The resulting gas flow of 985 m3 per hour is small comparedto the present total LNGS ventilation rate of 40,000 m3/h - which soon will be increasedconsiderably. For LAr the corresponding gas flow is slightly lower (see Appendix A).

41

Leakage of outer vessel: A leak in the outer vessel will destroy the insulation vacuumand enable the surrounding water to enter the space between inner and outer vessel. Waterwith its very high specific heat capacity of 4.19 kJ/(kgK) represents an efficient heater forthe LN; a drop of less than 4◦C in the temperature of the 660 m3 stored water is enoughto supply all the heat needed to evaporate the 46 m3 of LN contained in the cryostat.

An estimate of the resulting gas flow is difficult lacking detailed knowledge of theinvolved heat transfer. The heat transfer in an unsteady state involving phase changeshas been studied by putting a 1.5 in diameter copper cylinder at 70◦C into a bath of LNand measuring its change of temperature with time. As long as a bubble layer (boiling)was present, the heat transfer coefficient was found to be (170 ± 10) W/(m2K) (valuesbetween 600 to 900 W/(m2K) were found for the case that a copper cylinder of 80 Kwas immersed into water of 70◦C) [Bro 01]. While all these values are known to varywith radius (∝1/r) and length of the rod, a complete understanding is still pending andagreement with theoretical predictions is poor. In another approach, studies at the MPIHeidelberg considered a metal can filled with LN (15 cm diameter, 15 cm height) whichwas suddenly immersed into a water bath. Very strong convection in the water and verystrong boiling were observed as well as the formation of an about 1 cm thick ice layer afterabout 3’. It took 3’20” until about half of the LN amount was evaporated, and 6’40” for thesecond half. From these measurements preliminary values for the heat transfer coefficientsof about 55[47] W/(m2K) before[after] ice formation have been deduced. Adopting for themoment a value of 100 W/(m2K), the initial gas flow will be 22,000 m3 per hour whichmight be tolerable after the upgrade of the ventilation system. In this scenario all LNwould be evaporated after less than 3 hours. However, the presence of the many foils ofthe superinsulation will definitely diminish the heat transfer.

Leakage of inner vessel: A leak in the inner vessel will destroy the insulation vacuumand enable LN to enter the space between inner and outer container. The LN will beefficiently heated by the water surrounding the outer vessel with consequences similiar tothose discussed above. The maximum gas flow might be, in fact, somewhat smaller thanin above scenario due to the isolating effect of the huge number of gas bubbles which willbe generated in the space between inner and outer vessel. However, other than in caseof a leak in the outer shell, the prevailing amount of the gas cannot escape through thecryostat’s neck but must be draine off through the pumping pipe(s). Thus the volume ofthe insulation vacuum must be equipped with overpressure safety valves of the same typeas used for the actual cryostat.

Leakage of inner and outer vessel: Leaks in both the inner and outer vessel can leadto a mixture of LN and water which in certain cases [Arc 04] may lead to explosion-likeevaporation of nitrogen. Thus the water vessel needs to be equipped with a adequateoverpressure safety device.

If leaks develop simultaneously in both shells of the cryostat it is highly desirable toempty the water vessel such that the cryostat is no longer in contact with water. Estimates

42

show that this could be done in less than 10 minutes using pipes of still acceptable diameter.

6.1.3 Safety Valves

According to the ASME Code it is recommended to use two safety devices to preventbuildup of pressure in the inner vessel: (i) a spring-loaded safety valve set at approximately1.1 times the working pressure, and (ii) a safety head or rupture disk set to burst atapproximately 1.2 times the working pressure [Bar 85]. The size of the safety valve isdetermined by the ASME Code by

A = m · (R · T/M)1/2

C ·K · pmax

with C = [γ · (2/(γ + 1))γ+1γ−1 ]1/2

where A is the discharge area of valve, m the maximum gas flow rate through valve, R theuniversal gas constant (8314 J/K·kmol), T the absolute temperature of the gas at the inletto the valve, M the molecular weight of gas, K the discharge coefficient, pmax 1.1 times theset gauge pressure plus the atmospheric pressure, and γ = Cp/CV (1.4 for N2). AssumingT=273 K, K = 1 and a set pressure of 1.5 bar absolute, Table 5 shows the required size

Table 5: Size of safety valves with a K = 1 discharge coefficient for indicated mass flow ratesof nitrogen at T=273 K; the set gauge pressure is 1.5 bar absolute. The mass flow per hourin percent refers to a initial total cryogenic mass of 4·104 kg. Sizes have been calculatedaccording the ASME VIII (1st value) and AD2000 A2 (2nd value) prescriptions using thesoftware Si-Tech 2.2 [B&R 05]. For a more realistic discharge coefficient of K = 0.7, thequoted values have to be increased by 20%.

mass flow rate valve diameterkg/s %/h mm0.78 7 49.3 / 51.43.7 33 107.0 / 111.5

for several mass flow rates. The first line refers to the case of loss of insulation vacuum,the second line to an event in which the outer shell is broken such that the space betweeninner and outer vessel is immediately filled with water.

6.1.4 Corrosion

The operation of a copper cryostat in de-ionized water contained in a steel or stainlesssteel vessel has the potential hazard of serious corrosion effects. A study of this topic is improgress at the University of Milano.

43

7 GERDA Assembly at Lngs

As outlined in chapter 2, the Gerda experiment will consist of a cryostat immersed intoa water tank, and a superstructure including laboratory building and platform with apenthouse overlapping the water tank. The various components of Gerda are envisagedto be build in the following sequence:

1. water tank,

2. super-structure,

3. laboratory building,

4. penthouse.

A detailed plan for the complete assembly procedure is still missing, and present effortsfocus at which step the cryostat can be installed.

The various fabrication steps of the water tank are known, and in one scenario thecryostat could be inserted in the tank as soon as the roof and two rings of the water tank’swall have been welded together:

• build the water tank’s bottom plate,

• build the conical roof leaving a hole of 4.5 m in its center,

• weld the first sheet-metal plate of shell to the roof,

• lift up the roof,

• weld the second sheet-metal plate of the shell at the first one,

• insert the inner cryostat through the 4.5 m hole in the roof into the water tank,

• complete the construction of the water tank by repeated welding and lifting,

• solder last part of the roof,

• X-ray test all solderings,

• clean soldering (decappaggio and passivazione),

• clean water tank with ultraclean water.

While this procedure is viable, it contradicts the schedule outlined in the Proposal. There,it had been assumed that the water tank will be finished before the estimated delivery ofthe cryostat in January 2005 - the laboratory building was not yet explicitly mentioned.

In an alternative scenario, the cryostat would be placed into the water tank after step 3,i.e. after the completion of water tank, super-structure, laboratory building, and parts ofthe penthouse. This would help to avoid potentially large delays - the delivery date of the

44

highly non-standard cryostat is not yet ensured - and to minimize hardware modificationsif the cryostat, after its installation, would have to be removed from the water tank forrepair. This scenario would require an adequate flange in the water tank through whichthe cryostat could be inserted in either vertical, horizontal or inclined position as well asadequate lifting tools to move the cryostat within the water tank into its final position.

45

8 The LArGe-facility

The LArGe-facility of Gerda is located in the former Lens barrack underground in theinterferometer tunnel adjacent to Luna–II. The structure of the barrack, the air handlingsystem, as well as the Lens shielding system are currently being modified for their futureuse in the Gerda project.

The purpose of the LArGe-facility is to provide underground laboratory space, shieldedagainst radiation from cosmic rays, to minimize the production of cosmogenic radio-isotopes of the enriched germanium diodes and of the detector support structure. It alsoprovides the utilities for R&D on a novel background reduction technique using the anti-coincidence signal provided by the scintillation light of liquid argon as veto in the searchfor neutrinoless double beta decay.

To meet these objectives, the barrack is currently being modified and new laboratoryinfrastructures are under preparation. The start-up of the LArGe–facility is on the ’crit-ical path’ for the Gerda experiment, because the existing enriched diodes are required tobe modified and tested timely for Phase-I of the experiment.

8.1 LArGe infrastructures

Figure 8.1 displays the floor plan and dimensions of the LArGe–facility. The barrack willbe operated as a clean room with access through a personnel lock from the TIR tunnel.Fresh air enters the barrack through a ventilation system with temperature and humiditycontrol. The air inventory (approx. 300 m3) of the barrack is exchanged about once perhour. It is constantly circulated and dust particles are removed by hepa-filters. Highpurity working zones of class < 100 will be achieved in dedicated clean benches (4,5 inFigure 8.1) and by mobile hepa-filters. Chemical treatment of detector materials as e.g.surface etching is carried out in a fume hood (1) equipped with activated charcoal filtersretaining vapors. An evaporation system to produce thin metal layers for electrical contactswill be installed (not shown in figure). Final detector cleaning, refurbishment of electricalcontacts and mounting in their support structure will be carried out in a radon-free cleanbench (5). A detector test device (6) is connected to this clean bench. It serves to test thedetector performance after their refurbishment without exposing them to ambient air. Thedevice consists of a moderately shielded (liquid nitrogen (argon) dewar which is equippedwith a detector insertion system. Filling of the cryogenic liquids and blanketing are carriedout with the gas and cryogenic liquid distribution system (10). The LArGe device (12)for studying the background levels of the refurbished detectors will be installed in thenorthern part of the barrack. Further general infrastructures as washstand with deionizedwater system (2), mounting tables (3), storage shelfs (8), data acquisition system (7,13)will be installed. Technical details of the infrastructure are given below.

1. Fume hood with ventilation unit and activated charcoal filter. Exhaust air from filtercan be vented either to the outside of the barrack or into the barrack (circulation

46

13m

4,6m

15m

ChangeTür

Cur

tain

/Wal

l

clean room

clean room

1 2

3

4 5 6

7

98

14

15

13

12

11

10

Fels

Fels

Figure 21: Floor plan of the LArGe barrack. Explanations cf. text.

mode). L x W x H: 1800 mm, 2500 mm, 850 mm. Air flow 0–800m3/hour; 380 V,max. 16 A.

2. Washstand and water purification system for the production of deionized water (Milli-pore system and/or quartz distillation).

3. Mounting table stainless steel (movable),

4. Clean bench, vertical air stream (class 100, Airstream AVC-6A1), LxWxH: 1950 x750 x (1250+860) mm, 1930 m3/h, 230 V, 2.5 A, < 62 dBA.

5. Rn-reduced clean bench operated under nitrogen atmosphere in circulation mode,Optional, operational in exhaust mode as fume hood. LxWxH: 1600 x 800 x 3000mm, 380 V.

6. Nitrogen/argon dewar connected to 5. Volume: 70 liters, max. operational overpressure: 1.5 bar; height: 1100 mm, diameter: 550 mm (plus wheels).

7. Data acquisition system for 6.

8. Storage shelfs for laboratory material and enriched detectors with dewars (Igex,HdM detectors prior to refurbishment).

47

9. Desk workspace with PC.

10. Liquid/gaseous nitrogen/argon handling system for (re-)filling, emptying and gasblanket for supply for 6 and 12.

11. Access ladder to top of 12.

12. LArGe system

13. LArGe data acquisition (14 bit SIS flash ADC, controlled by Intel PC operatedunder Linux) and slow control.

14. Liquid argon storage approx. 6 m3

15. Purification system for water, oxygen and radon removal from liquid argon.

8.2 The LArGe system

A vacuum insulated copper dewar of 900/1000 mm inner/outer diameter and 2050 mmheight is housed in a graded shielding system of 200 mm PE, 230 mm steel, 100 mm low-activity lead and 150 mm electrolytic copper. A gas-tight lock mounted on top of the shieldserves as the access port for detector insertion. The system is hermetically gas tight andkept under a overpressure of argon atmosphere of approximately 10 mbar to avoid radoncontaminations.

Filling and emptying of the cryogenic liquid inventory (1000 liter of liquefied argon)occurs via a dedicated liquid/gas handling system. Necessary safety related instrumenta-tions as well as burst disks and ventilation tubes are integrated into the system. Liquidargon for initial filling and re-filling during operations is stored in a storage dewar (6 m3)external to the barrack. Adjacent to the liquid argon storage is a liquid argon purificationsystem based on the LTA technique (low temperature adsorption on high purity charcoal)developed for Borexino. Optional, an oxygen removal cartridge (Oxysorb) is insertedupstream from the LTA to prevent emanated radon to reach the LArGe dewar.

A detailed description of the LArGe system is given in a separate report.

8.3 Time schedule

The modification of the barrack structure – ie the removal of walls, the refurbishment of thefloor and coating with epoxy resin, enlargement of northern barrack part, modification ofthe air ventilation and hepa system – started in February 2005 and will be completed earlyApril. Installation of the graded shielding and the laboratory infrastructures is scheduledfor April and early May. Work with detectors in a clean environment will commence inMay. Completion of the installation of the full LArGe system is targeted for summer2005.

48

Appendix A

Table 6: Physical properties of nitrogen and argon

Property Argon Nitrogen Unit

Boiling point at 101.3 kPa (1 bar) -185.5 -195.8 ◦C87.28 77.36 K

Density of liquid at boiling point 1394 807 kg/m3

Specific heat cp at boiling point 1.14 2.05 kJ/kg·KThermal conductivity λ at boiling point 123 140 mW/m·KDielectric constant εr at boiling point 1.52 1.43Latent heat of evaporation 161.9 199.3 kJ/kgAmount of gas (20◦C, 0.1 MPa) 841 693 m3

produced by 1 m3 of liquidDensity at 20◦C compared to 1.4 1.0density of airRatio of enthalpy of vapor at 20◦C 0.7 1.14and latent heat of evaporation

49

References

[AD 2000] AD 2000-Regelwerk, Verband der Technischen Uberwachungsvereine e.V., Es-sen, 2004, Karl Heymanns Verlag KG, Koln.

[AAe 05] Austrian Aerospace, datasheet 2005.

[Arc 04] U. Archakositt, S. Nilsuwankosit, and T. Sumitra, J. Nucl. Sci. Tech.41 (2004),432-439.

[Bar 85] R.F. Barron, Cryogenic Systems, Oxford University Press, 1985,ISBN 0-19-503567-4.

[B&R 05] Bopp & Reuther Sicherheits- und Regelarmaturen GmbH, Mannheim,http://www.sr.boppureuther.com.

[Bro 01] L. Brown, S. Carothers, M. Monyok, and Z. Smith, ‘Incorporating PhaseChanges with Unsteady State Heat Transfer - S01’,http://rothfus.cheme.cmu.edu/tlab/ussht/projects/index.htm.

[Fes 02] J. Fesmire, S. Augustynowicz, C. Darve, ‘Performance characterization of per-forated multilayer insulation blankets’ ICEC 19, Grenoble, July 2002, http://tdserver1.fnal.gov/nicol/lhc−irq−cryostat/ch−darve/public/publication.htm.

[Ger 04] GERDA, ‘The GERmanium Detector Array for the search of neutrinoless dou-ble beta decay in Ge-76 at Lngs’, The GERDA Collaboration, Proposal toLNGS P38/04, September 2004.

[Pro 04] pro-beam AG, 39288 Burg.

[Tim 89] K.D. Timmerhaus and Th.M. Flynn, Cryogenic Process Engineering, PlenumPress, 1989, ISBN 0-306-43283-8.

50